Tian Suqing, Wang Cuiying, Zhang Ruiping, Dai Zhuojie, Jia Lecheng, Zhang Wei, Wang Junjie, Liu Yinglong
Department of Radiation Oncology, Peking University Third Hospital, Beijing, China.
Department of Oncology, Hainan Third People's Hospital, Sanya, China.
Front Oncol. 2022 Apr 14;12:856346. doi: 10.3389/fonc.2022.856346. eCollection 2022.
Glioblastoma is the most common primary malignant brain tumor in adults and can be treated with radiation therapy. However, tumor target contouring for head radiation therapy is labor-intensive and highly dependent on the experience of the radiation oncologist. Recently, autosegmentation of the tumor target has been playing an increasingly important role in the development of radiotherapy plans. Therefore, we established a deep learning model and improved its performance in autosegmenting and contouring the primary gross tumor volume (GTV) of glioblastomas through transfer learning.
The preoperative MRI data of 20 patients with glioblastomas were collected from our department (ST) and split into a training set and testing set. We fine-tuned a deep learning model for autosegmentation of the hippocampus on separate MRI scans (RZ) through transfer learning and trained this deep learning model directly using the training set. Finally, we evaluated the performance of both trained models in autosegmenting glioblastomas using the testing set.
The fine-tuned model converged within 20 epochs, compared to over 50 epochs for the model trained directly by the same training set, and demonstrated better autosegmentation performance [Dice similarity coefficient (DSC) 0.9404 ± 0.0117, 95% Hausdorff distance (95HD) 1.8107 mm ±0.3964, average surface distance (ASD) 0.6003 mm ±0.1287mm] than the model trained directly (DSC 0.9158±0.0178, 95HD 2.5761 mm ± 0.5365mm, ASD 0.7579 mm ± 0.1468mm) with the same test set. The DSC, 95HD, and ASD values of the two models were significantly different (<0.05).
A model developed with semisupervised transfer learning and trained on independent data achieved good performance in autosegmenting glioblastoma. The autosegmented volume of glioblastomas is sufficiently accurate for radiotherapy treatment, which could have a positive impact on tumor control and patient survival.
胶质母细胞瘤是成人中最常见的原发性恶性脑肿瘤，可采用放射治疗。然而，头部放射治疗的肿瘤靶区勾画工作强度大，且高度依赖放射肿瘤学家的经验。近年来，肿瘤靶区的自动分割在放射治疗计划的制定中发挥着越来越重要的作用。因此，我们建立了一个深度学习模型，并通过迁移学习提高其在自动分割和勾画胶质母细胞瘤原发性大体肿瘤体积（GTV）方面的性能。
收集了本部门（ST）20例胶质母细胞瘤患者的术前MRI数据，并将其分为训练集和测试集。我们通过迁移学习在单独的MRI扫描（RZ）上对用于海马自动分割的深度学习模型进行微调，并使用训练集直接训练该深度学习模型。最后，我们使用测试集评估了两个训练模型在自动分割胶质母细胞瘤方面的性能。
微调后的模型在20个epoch内收敛，而使用相同训练集直接训练的模型则需要超过50个epoch，并且在自动分割性能方面[骰子相似系数（DSC）0.9404±0.0117，95%豪斯多夫距离（95HD）1.8107mm±0.3964，平均表面距离（ASD）0.6003mm±0.1287mm]优于使用相同测试集直接训练的模型（DSC 0.9158±0.0178，95HD 2.5761mm±0.5365mm，ASD 0.7579mm±0.1468mm）。两个模型的DSC、95HD和ASD值有显著差异（<0.05）。
通过半监督迁移学习开发并在独立数据上训练的模型在自动分割胶质母细胞瘤方面取得了良好的性能。胶质母细胞瘤的自动分割体积对于放射治疗足够准确，这可能对肿瘤控制和患者生存产生积极影响。
